Free tool

Loudness targets reference

Delivery targets for podcasts, audiobooks, streaming and broadcast, checked against each publisher's own documentation. Where a platform publishes no number, the table says so.

17 of 17 entries shown
Loudness delivery targets by standard and platform
Standard or platformIntegrated targetTolerancePeak ceilingOther limitsNotes and source
Apple PodcastsPodcast −16 LKFS ±1 dB True peak ≤ −1 dBFS (BS.1770-5) AAC preferred, MP3 accepted. Mono 64–128 kbps, stereo 128–256 kbps at 44.1 or 48 kHz. Official recommendation. Apple Podcasts for Creators
Mono podcast files (convention)Podcast −19 LUFS – – Plays like −16 LUFS stereo when the mono file is played on two speakers (about 3 LU louder). Popularised by Google's Assistant loudness guidance, now deprecated. A common convention, not an AES recommendation. Google Assistant audio loudness (deprecated)
Spotify podcastsPodcast None published – – Uploads in MP3, M4A or WAV, mono or stereo. Spotify has not published a loudness target for podcast episodes. Spotify for Creators
ACX / Audible audiobooksAudiobook No LUFS target. RMS −23 to −18 dB Range Peak ≤ −3 dB (type not specified; ACX Check measures sample peak) Noise floor ≤ −60 dB RMS. MP3 192 kbps or higher, CBR, 44.1 kHz, no joint stereo. Max 120 minutes per file. 1–5 s of room tone at head and tail. Official requirements. ACX Audio Submission Requirements
Spotify music (playback normalization)Streaming −14 LUFS (Normal, default). Loud −11, Quiet −19 Not stated Gain increases leave 1 dB of true-peak headroom Spotify advises mastering at −14 LUFS with true peak below −1 dBTP, or below −2 dBTP if the master is louder than −14 LUFS. Applies to music playback, not podcast uploads. Spotify for Artists: loudness normalization
Spotify audio adsStreaming −16 LUFS ±1.5 LU Not stated 192–320 kbps, 44.1 kHz, stereo. Ad creatives only. Spotify Advertising audio ad specs
YouTubeVideo No official target – – Recommended upload audio: AAC-LC or Opus at 48 kHz. "Stable volume" is on by default for viewers. −14 LUFS is commonly reported, not an official specification. YouTube recommended upload encoding
Amazon Music, Deezer, TidalStreaming None published – – Loudness figures circulating online for these services are not published by them. No official specification found.
Amazon Alexa skills and flash briefingsVoice −14 LUFS average Rejected below −19 or above −9 LUFS True peak ≤ −2 dBFS – Official developer guidance. Amazon Alexa: normalizing loudness
AES TD1008 (2021), internet streamingStreaming Speech −18 LUFS. Music: track −16, album −14 Upper tolerance +1 LU for speech (dialogue-measured) ≤ −1 dBTP at the codec input Keep integrated loudness above −20 LUFS. Notes that a mono file played on two channels measures 3 LU higher. Superseded by AES77-2023, whose figures we have not verified. AES TD1008
AES TD1004 (2015), streamingStreaming −20 to −16 LUFS – ≤ −1.0 dBTP Short-form: maximum short-term loudness no more than 5 LU above the target. Superseded by TD1008. AES TD1004
EBU R 128 (V5, 2023)Broadcast −23.0 LUFS ±1.0 LU only where the target is not practically achievable (e.g. live). ±0.2 LU for QC measurement ≤ −1 dBTP No maximum loudness range; LRA is not recommended for programmes under 1 minute. European TV and radio production. EBU R 128
EBU R 128 s1 (short-form, adverts)Broadcast −23.0 LUFS ±0.2 LU (QC) ≤ −1 dBTP Maximum short-term loudness ≤ −18.0 LUFS. Content up to about 2 minutes. EBU R 128 s1
EBU R 128 s2 (streaming)Streaming −23.0 LUFS; interim distribution level −20.0 to −16.0 LUFS – ≤ −1 dBTP (production) Broadcasters are asked to move to −23 LUFS once playback devices allow. Broadcaster streaming services. EBU R 128 s2
ATSC A/85 (US television)Broadcast −24 LKFS (dialogue anchor for long-form) About ±2 dB measurement tolerance; do not target the edges Below −2 dBTP Streaming annex: −23 to −27 LKFS. Current revision A/85:2026-07. US FCC rules still cite A/85:2013 (same targets). ATSC A/85:2026-07
Free TV Australia OP-59Broadcast −24 LKFS ±1 LU ≤ −2 dBTP – Australian free-to-air television. Free TV Australia OP-59
Netflix (Sound Mix Specifications v1.6)Video −27 LKFS, dialogue-gated ±2 LU ≤ −2 dBTP −24 LKFS ±2 LU when less than 15% of the programme is dialogue. Read from an archived copy; the live specification may have changed.

Values checked against each publisher's own documentation on 23 October 2026. "LUFS" and "LKFS" are the same unit. Platforms change their guidance from time to time, so confirm the current specification before an important delivery.

LUFS, RMS and true peak: what each one measures

LUFS (loudness)

Loudness Units relative to Full Scale, defined in ITU-R BS.1770-5. The signal is K-weighted so that frequencies count roughly as much as we hear them, averaged in 400 ms blocks, and gated so silence and very quiet passages do not drag the average down. Because it tracks perceived loudness, it is what streaming platforms and broadcasters normalize to. Measure any file with the LUFS meter.

RMS (average level)

The root mean square of the samples: a plain average of signal power with no frequency weighting and no gating. Pauses count, and a bass-heavy recording reads higher than it sounds. ACX still specifies audiobooks in RMS (−23 to −18 dB), so use the ACX check for those rather than a LUFS figure. There is no official conversion between the two.

True peak (dBTP)

The highest point of the waveform after it is reconstructed between samples, measured with at least four times oversampling. It can be higher than the highest sample (the sample peak, in dBFS), and it is what MP3 and AAC encoders and playback devices actually have to handle. That is why delivery specifications set the ceiling in dBTP.

Putting it into practice

Choose the row that matches where your audio is going, measure your file, and adjust the gain: the gain calculator shows how much and whether a limiter is needed. LufsReady's delivery presets master to these targets automatically, and the pricing page explains how minutes work.

Frequently asked questions

Which target should I use for a podcast?
Apple Podcasts recommends −16 LKFS ±1 dB with a true peak no higher than −1 dBFS, and that is the safest general choice for stereo episodes. Mono files are commonly mastered to −19 LUFS so they play at the same level on two speakers. AES TD1008 suggests −18 LUFS for speech streams. Spotify has published no podcast target.
Is −14 LUFS the YouTube or Spotify standard?
Spotify normalizes music playback to −14 LUFS by default and says so in its documentation. YouTube publishes no loudness target: −14 LUFS is a figure commonly observed by third parties, not an official specification, and Spotify does not say whether music normalization applies to podcasts.
What is the difference between LUFS and LKFS?
None. Both name the same BS.1770 loudness measurement: LUFS is the EBU term, LKFS the ITU and ATSC term. A difference of 1 LU equals 1 dB.
Why do so many specifications ask for −1 or −2 dBTP?
Lossy encoders such as MP3 and AAC, and some playback devices, can create peaks higher than those in the original file. Keeping the true peak a decibel or two below full scale leaves room for that without audible clipping.

Master it to spec automatically

Upload a recording, pick Apple Podcasts, ACX, EBU R 128 or your own target, and download a file that meets it, with a before-and-after report. New accounts get 60 free minutes after confirming their email.